Fusion of deterministically generated photonic graph states.

Entanglement has evolved from an enigmatic concept of quantum physics to a key ingredient of quantum technology. It explains correlations between measurement outcomes that contradict classical physics and has been widely explored with small sets of individual qubits. Multi-partite entangled states build up in gate-based quantum-computing protocols and-from a broader perspective-were proposed as the main resource for measurement-based quantum-information processing1,2. The latter requires the ex-ante generation of a multi-qubit entangled state described by a graph3-6. Small graph states such as Bell or linear cluster states have been produced with photons7-16, but the proposed quantum-computing and quantum-networking applications require fusion of such states into larger and more powerful states in a programmable fashion17-21. Here we achieve this goal by using an optical resonator22 containing two individually addressable atoms23,24. Ring25 and tree26 graph states with up to eight qubits, with the names reflecting the entanglement topology, are efficiently fused from the photonic states emitted by the individual atoms. The fusion process itself uses a cavity-assisted gate between the two atoms. Our technique is, in principle, scalable to even larger numbers of qubits and is the decisive step towards, for instance, a memory-less quantum repeater in a future quantum internet27-29.

Efficient generation of entangled multiphoton graph states from a single atom.

The central technological appeal of quantum science resides in exploiting quantum effects, such as entanglement, for a variety of applications, including computing, communication and sensing1. The overarching challenge in these fields is to address, control and protect systems of many qubits against decoherence2. Against this backdrop, optical photons, naturally robust and easy to manipulate, represent ideal qubit carriers. However, the most successful technique so far for creating photonic entanglement3 is inherently probabilistic and, therefore, subject to severe scalability limitations. Here we report the implementation of a deterministic protocol4-6 for the creation of photonic entanglement with a single memory atom in a cavity7. We interleave controlled single-photon emissions with tailored atomic qubit rotations to efficiently grow Greenberger-Horne-Zeilinger (GHZ) states8 of up to 14 photons and linear cluster states9 of up to 12 photons with a fidelity lower bounded by 76(6)% and 56(4)%, respectively. Thanks to a source-to-detection efficiency of 43.18(7)% per photon, we measure these large states about once every minute, which is orders of magnitude faster than in any previous experiment3,10-13. In the future, this rate could be increased even further, the scheme could be extended to two atoms in a cavity14,15 or several sources could be quantum mechanically coupled16, to generate higher-dimensional cluster states17. Overcoming the limitations encountered by probabilistic schemes for photonic entanglement generation, our results may offer a way towards scalable measurement-based quantum computation18,19 and communication20,21.

The Image Biomarker Standardization Initiative: Standardized Convolutional Filters for Reproducible Radiomics and Enhanced Clinical Insights.

Filters are commonly used to enhance specific structures and patterns in images, such as vessels or peritumoral regions, to enable clinical insights beyond the visible image using radiomics. However, their lack of standardization restricts reproducibility and clinical translation of radiomics decision support tools. In this special report, teams of researchers who developed radiomics software participated in a three-phase study (September 2020 to December 2022) to establish a standardized set of filters. The first two phases focused on finding reference filtered images and reference feature values for commonly used convolutional filters: mean, Laplacian of Gaussian, Laws and Gabor kernels, separable and nonseparable wavelets (including decomposed forms), and Riesz transformations. In the first phase, 15 teams used digital phantoms to establish 33 reference filtered images of 36 filter configurations. In phase 2, 11 teams used a chest CT image to derive reference values for 323 of 396 features computed from filtered images using 22 filter and image processing configurations. Reference filtered images and feature values for Riesz transformations were not established. Reproducibility of standardized convolutional filters was validated on a public data set of multimodal imaging (CT, fluorodeoxyglucose PET, and T1-weighted MRI) in 51 patients with soft-tissue sarcoma. At validation, reproducibility of 486 features computed from filtered images using nine configurations × three imaging modalities was assessed using the lower bounds of 95% CIs of intraclass correlation coefficients. Out of 486 features, 458 were found to be reproducible across nine teams with lower bounds of 95% CIs of intraclass correlation coefficients greater than 0.75. In conclusion, eight filter types were standardized with reference filtered images and reference feature values for verifying and calibrating radiomics software packages. A web-based tool is available for compliance checking.

Adverse radiation effect versus tumor progression following stereotactic radiosurgery for brain metastases: Implications of radiologic uncertainty.

BACKGROUND: Adverse radiation effect (ARE) following stereotactic radiosurgery (SRS) for brain metastases is challenging to distinguish from tumor progression. This study characterizes the clinical implications of radiologic uncertainty (RU). METHODS: Cases reviewed retrospectively at a single-institutional, multi-disciplinary SRS Tumor Board between 2015-2022 for RU following SRS were identified. Treatment history, diagnostic or therapeutic interventions performed upon RU resolution, and development of neurologic deficits surrounding intervention were obtained from the medical record. Differences in lesion volume and maximum diameter at RU onset versus resolution were compared with paired t-tests. Median time from RU onset to resolution was estimated using the Kaplan-Meier method. Univariate and multivariate associations between clinical characteristics and time to RU resolution were assessed with Cox proportional-hazards regression. RESULTS: Among 128 lesions with RU, 23.5% had undergone ≥ 2 courses of radiation. Median maximum diameter (20 vs. 16 mm, p < 0.001) and volume (2.7 vs. 1.5 cc, p < 0.001) were larger upon RU resolution versus onset. RU resolution took > 6 and > 12 months in 25% and 7% of cases, respectively. Higher total EQD2 prior to RU onset (HR = 0.45, p = 0.03) and use of MR perfusion (HR = 0.56, p = 0.001) correlated with shorter time to resolution; larger volume (HR = 1.05, p = 0.006) portended longer time to resolution. Most lesions (57%) were diagnosed as ARE. Most patients (58%) underwent an intervention upon RU resolution; of these, 38% developed a neurologic deficit surrounding intervention. CONCLUSIONS: RU resolution took > 6 months in > 25% of cases. RU may lead to suboptimal outcomes and symptom burden. Improved characterization of post-SRS RU is needed.

Puzzling out graphic codes.

This response takes advantage of the diverse and wide-ranging series of commentaries to clarify some aspects of the target article, and flesh out other aspects. My central point is a plea to take graphic codes seriously as codes, rather than as a kind of visual art or as a byproduct of spoken language; only in this way can the puzzle of ideography be identified and solved. In this perspective, I argue that graphic codes do not derive their expressive power from iconicity alone (unlike visual arts), and I clarify the peculiar relationship that ties writing to spoken language. I then discuss three possible solutions to the puzzle of ideography. I argue that a learning account still cannot explain why ideographies fail to evolve, even if we emancipate the learning account from the version that Liberman put forward; I develop my preferred solution, the "standardization account," and contrast it with a third solution suggested by some commentaries, which says that ideographies do not evolve because they would make communication too costly. I consider, by way of conclusion, the consequences of these views for the future evolution of ideography.

Association of mental health diagnosis with race and all-cause mortality after a cancer diagnosis: Large-scale analysis of electronic health record data.

BACKGROUND: Disparity in mental health care among cancer patients remains understudied. METHODS: A large, retrospective, single tertiary-care institution cohort study was conducted based on deidentified electronic health record data of 54,852 adult cancer patients without prior mental health diagnosis (MHD) diagnosed at the University of California, San Francisco between January 2012 and September 2019. The exposure of interest was early-onset MHD with or without psychotropic medication (PM) within 12 months of cancer diagnosis and primary outcome was all-cause mortality. RESULTS: There were 8.2% of patients who received a new MHD at a median of 197 days (interquartile range, 61-553) after incident cancer diagnosis; 31.0% received a PM prescription; and 3.7% a mental health-related visit (MHRV). There were 62.6% of patients who were non-Hispanic White (NHW), 10.8% were Asian, 9.8% were Hispanic, and 3.8% were Black. Compared with NHWs, minority cancer patients had reduced adjusted odds of MHDs, PM prescriptions, and MHRVs, particularly for generalized anxiety (Asian odds ratio [OR], 0.66, 95% CI, 0.55-0.78; Black OR, 0.60, 95% CI, 0.45-0.79; Hispanic OR, 0.72, 95% CI, 0.61-0.85) and selective serotonin-reuptake inhibitors (Asian OR, 0.43, 95% CI, 0.37-0.50; Black OR, 0.51, 95% CI, 0.40-0.61; Hispanic OR, 0.79, 95% CI, 0.70-0.89). New early MHD with PM was associated with elevated all-cause mortality (12-24 months: hazard ratio [HR], 1.43, 95% CI, 1.25-1.64) that waned by 24 to 36 months (HR, 1.18, 95% CI, 0.95-1.45). CONCLUSIONS: New mental health diagnosis with PM was a marker of early mortality among cancer patients. Minority cancer patients were less likely to receive documentation of MHDs or treatment, which may represent missed opportunities to identify and treat cancer-related mental health conditions.

The puzzle of ideography.

An ideography is a general-purpose code made of pictures that do not encode language, which can be used autonomously - not just as a mnemonic prop - to encode information on a broad range of topics. Why are viable ideographies so hard to find? I contend that self-sufficient graphic codes need to be narrowly specialized. Writing systems are only an apparent exception: At their core, they are notations of a spoken language. Even if they also encode nonlinguistic information, they are useless to someone who lacks linguistic competence in the encoded language or a related one. The versatility of writing is thus vicarious: Writing borrows it from spoken language. Why is it so difficult to build a fully generalist graphic code? The most widespread answer points to a learnability problem. We possess specialized cognitive resources for learning spoken language, but lack them for graphic codes. I argue in favor of a different account: What is difficult about graphic codes is not so much learning or teaching them as getting every user to learn and teach the same code. This standardization problem does not affect spoken or signed languages as much. Those are based on cheap and transient signals, allowing for easy online repairing of miscommunication, and require face-to-face interactions where the advantages of common ground are maximized. Graphic codes lack these advantages, which makes them smaller in size and more specialized.

Quantum Teleportation between Remote Qubit Memories with Only a Single Photon as a Resource.

Quantum teleportation enables the deterministic exchange of qubits via lossy channels. While it is commonly believed that unconditional teleportation requires a preshared entangled qubit pair, here we demonstrate a protocol that is in principle unconditional and requires only a single photon as an ex-ante prepared resource. The photon successively interacts, first, with the receiver and then with the sender qubit memory. Its detection, followed by classical communication, heralds a successful teleportation. We teleport six mutually unbiased qubit states with average fidelity F[over ¯]=(88.3±1.3)% at a rate of 6 Hz over 60 m.

Detecting an Itinerant Optical Photon Twice without Destroying It.

Nondestructive quantum measurements are central for quantum physics applications ranging from quantum sensing to quantum computing and quantum communication. Employing the toolbox of cavity quantum electrodynamics, we here concatenate two identical nondestructive photon detectors to repeatedly detect and track a single photon propagating through a 60 m long optical fiber. By demonstrating that the combined signal-to-noise ratio of the two detectors surpasses each single one by about 2 orders of magnitude, we experimentally verify a key practical benefit of cascaded nondemolition detectors compared to conventional absorbing devices.

Feasibility and Significance of Dose Adaptation via Linear Couch Translations to Correct for Rotational Shifts During Frameless Brain Radiosurgery with the Gamma Knife Icon™.

OBJECTIVE: The present study aimed to examine the technical feasibility and effectiveness of adapting the radiation dose distributions with three-dimensional (3D) linear couch translations in contrast to full six-dimensional couch maneuvers to correct for rotational shifts during frameless radiosurgical treatment with the Gamma Knife Icon™ (Elekta AB; Stockholm, Sweden). METHODS: The original magnetic resonance images used for radiosurgery treatment planning (15 targets) were digitally processed to simulate rotational shifts of ±1, ±2, ±3, ±5, and ±10 degrees in the transverse plane and imported back into Leksell GammaPlan® (Elekta AB), creating "uncorrected" treatment plans. In addition, geometrically optimized 3D translation shifts were consequently applied to each isocenter in all "uncorrected" treatment plans to account for systematically introduced rotational shifts and to produce "corrected" treatment plans. The differences in the dose distribution between the original treatment plans and the "uncorrected" and "corrected" treatment plans were calculated and compared at each rotational shift position. RESULTS: The "uncorrected" treatment plans resulted in a significant deterioration in target coverage (by 8-72%) and selectivity (by 2-42%), with some targets being missed completely with rotations of ±3 or more degrees. In contrast, in all "corrected" treatment plans, the average decreases in target coverage and selectivity were only 1% (maximum values 4-5%). CONCLUSION: Applications of 3D linear couch translations successfully overcome gross uncertainties in dose distributions caused by up to ±10 degrees of rotational shifts in a target. As a result, rapid dose adaptation with 3D couch translations is unique and effective for frameless radiosurgery with the Gamma Knife Icon™.

The Image Biomarker Standardization Initiative: Standardized Quantitative Radiomics for High-Throughput Image-based Phenotyping.

Background Radiomic features may quantify characteristics present in medical imaging. However, the lack of standardized definitions and validated reference values have hampered clinical use. Purpose To standardize a set of 174 radiomic features. Materials and Methods Radiomic features were assessed in three phases. In phase I, 487 features were derived from the basic set of 174 features. Twenty-five research teams with unique radiomics software implementations computed feature values directly from a digital phantom, without any additional image processing. In phase II, 15 teams computed values for 1347 derived features using a CT image of a patient with lung cancer and predefined image processing configurations. In both phases, consensus among the teams on the validity of tentative reference values was measured through the frequency of the modal value and classified as follows: less than three matches, weak; three to five matches, moderate; six to nine matches, strong; 10 or more matches, very strong. In the final phase (phase III), a public data set of multimodality images (CT, fluorine 18 fluorodeoxyglucose PET, and T1-weighted MRI) from 51 patients with soft-tissue sarcoma was used to prospectively assess reproducibility of standardized features. Results Consensus on reference values was initially weak for 232 of 302 features (76.8%) at phase I and 703 of 1075 features (65.4%) at phase II. At the final iteration, weak consensus remained for only two of 487 features (0.4%) at phase I and 19 of 1347 features (1.4%) at phase II. Strong or better consensus was achieved for 463 of 487 features (95.1%) at phase I and 1220 of 1347 features (90.6%) at phase II. Overall, 169 of 174 features were standardized in the first two phases. In the final validation phase (phase III), most of the 169 standardized features could be excellently reproduced (166 with CT; 164 with PET; and 164 with MRI). Conclusion A set of 169 radiomics features was standardized, which enabled verification and calibration of different radiomics software. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Kuhl and Truhn in this issue.

Durable complete remission following anti-EGFR antibodies in recurrent metastatic colorectal cancer.

In this case report, we describe a patient who remains in complete remission two years after the discontinuation of anti-EGFR monotherapy as a third-line treatment, accompanied by persistent severe hypomagnesemia. A 45-year-old Caucasian woman with mCRC started chemotherapy with weekly cetuximab. After ten months of treatment, the therapy was stopped because the patient had persistent grade III hypomagnesemia despite amiloride, oral, and intravenous magnesium. A month later, the patient was switched to panitumumab 6 mg/kg every two weeks for four additional months to avoid weekly visits to the clinic. Following discontinuation of panitumumab, PET scans remain negative to this day, two years after anti-EGFR therapy discontinuation. No factor has been identified to explain the complete and sustained response experienced by this patient. Hypomagnesemia is a common adverse effect of anti-EGFR therapy that can lead to treatment interruption and discontinuation if severe. This case highlights the importance of pursuing anti-EGFR therapy when a response is observed in spite of severe hypomagnesemia. It also provides preliminary information that anti-EGFR therapy could be stopped after a complete response is obtained.

Birth of the cool: a two-centuries decline in emotional expression in Anglophone fiction.

The presence of emotional words and content in stories has been shown to enhance a story's memorability, and its cultural success. Yet, recent cultural trends run in the opposite direction. Using the Google Books corpus, coupled with two metadata-rich corpora of Anglophone fiction books, we show a decrease in emotionality in English-speaking literature starting plausibly in the nineteenth century. We show that this decrease cannot be explained by changes unrelated to emotionality (such as demographic dynamics concerning age or gender balance, changes in vocabulary richness, or changes in the prevalence of literary genres), and that, in our three corpora, the decrease is driven almost entirely by a decline in the proportion of positive emotion-related words, while the frequency of negative emotion-related words shows little if any decline. Consistently with previous studies, we also find a link between ageing and negative emotionality at the individual level.

Optimistic expectations about communication explain children's difficulties in hiding, lying, and mistrusting liars.

We suggest that preschoolers' frequent obliviousness to the risks and opportunities of deception comes from a trusting stance supporting verbal communication. Three studies (N = 125) confirm this hypothesis. Three-year-olds can hide information from others (Study 1) and they can lie (Study 2) in simple settings. Yet when one introduces the possibility of informing others in the very same settings, three-year-olds tend to be honest (Studies 1 and 2). Similarly, four-year-olds, though capable of treating assertions as false, trust deceptive informants (Study 3). We suggest that children's reduced sensitivity to the opportunities of lying, and to the risks of being lied to might help explain their difficulties on standard false belief tasks.

High-efficiency WSi superconducting nanowire single-photon detectors for quantum state engineering in the near infrared.

We report on high-efficiency superconducting nanowire single-photon detectors based on amorphous tungsten silicide and optimized at 1064 nm. At an operating temperature of 1.8 K, we demonstrated a 93% system detection efficiency at this wavelength with a dark noise of a few counts per second. Combined with cavity-enhanced spontaneous parametric downconversion, this fiber-coupled detector enabled us to generate narrowband single photons with a heralding efficiency greater than 90% and a high spectral brightness of 0.6×104 photons/(s·mW·MHz). Beyond single-photon generation at large rate, such high-efficiency detectors open the path to efficient multiple-photon heralding and complex quantum state engineering.

Assessment of image quality and dose calculation accuracy on kV CBCT, MV CBCT, and MV CT images for urgent palliative radiotherapy treatments.

A clinical workflow was developed for urgent palliative radiotherapy treatments that integrates patient simulation, planning, quality assurance, and treatment in one 30-minute session. This has been successfully tested and implemented clinically on a linac with MV CBCT capabilities. To make this approach available to all clin-ics equipped with common imaging systems, dose calculation accuracy based on treatment sites was assessed for other imaging units. We evaluated the feasibility of palliative treatment planning using on-board imaging with respect to image quality and technical challenges. The purpose was to test multiple systems using their commercial setup, disregarding any additional in-house development. kV CT, kV CBCT, MV CBCT, and MV CT images of water and anthropomorphic phantoms were acquired on five different imaging units (Philips MX8000 CT Scanner, and Varian TrueBeam, Elekta VersaHD, Siemens Artiste, and Accuray Tomotherapy linacs). Image quality (noise, contrast, uniformity, spatial resolution) was evaluated and compared across all machines. Using individual image value to density calibrations, dose calculation accuracies for simple treatment plans were assessed for the same phantom images. Finally, image artifacts on clinical patient images were evaluated and compared among the machines. Image contrast to visualize bony anatomy was sufficient on all machines. Despite a high noise level and low contrast, MV CT images provided the most accurate treatment plans relative to kV CT-based planning. Spatial resolution was poorest for MV CBCT, but did not limit the visualization of small anatomical structures. A comparison of treatment plans showed that monitor units calculated based on a prescription point were within 5% difference relative to kV CT-based plans for all machines and all studied treatment sites (brain, neck, and pelvis). Local dose differences > 5% were found near the phantom edges. The gamma index for 3%/3 mm criteria was ≥ 95% in most cases. Best dose calculation results were obtained when the treatment isocenter was near the image isocenter for all machines. A large field of view and immediate image export to the treatment planning system were essential for a smooth workflow and were not provided on all devices. Based on this phantom study, image quality of the studied kV CBCT, MV CBCT, and MV CT on-board imaging devices was sufficient for treatment planning in all tested cases. Treatment plans provided dose calculation accuracies within an acceptable range for simple, urgently planned palliative treatments. However, dose calculation accuracy was compromised towards the edges of an image. Feasibility for clinical implementation should be assessed separately and may be complicated by machine specific features. Image artifacts in patient images and the effect on dose calculation accuracy should be assessed in a separate, machine-specific study.

The disunity of cultural group selection.

I argue that demographic selection, migration, and cultural diffusion, three mechanisms of institutional change, have little in common. Two of these lack the key features associated with group selection: they do not present us with group-level selection pressures counteracting individual-level ones, need not produce behavioral altruism, and do not require competition between groups whose members cooperate preferentially with one another. Cultural norms vary, change, and influence cooperation; but that is not group selection.

Feasibility of MV CBCT-based treatment planning for urgent radiation therapy: dosimetric accuracy of MV CBCT-based dose calculations.

Unlike scheduled radiotherapy treatments, treatment planning time and resources are limited for emergency treatments. Consequently, plans are often simple 2D image-based treatments that lag behind technical capabilities available for nonurgent radiotherapy. We have developed a novel integrated urgent workflow that uses onboard MV CBCT imaging for patient simulation to improve planning accuracy and reduce the total time for urgent treatments. This study evaluates both MV CBCT dose planning accuracy and novel urgent workflow feasibility for a variety of anatomic sites. We sought to limit local mean dose differences to less than 5% compared to conventional CT simulation. To improve dose calculation accuracy, we created separate Hounsfield unit-to-density calibration curves for regular and extended field-of-view (FOV) MV CBCTs. We evaluated dose calculation accuracy on phantoms and four clinical anatomical sites (brain, thorax/spine, pelvis, and extremities). Plans were created for each case and dose was calculated on both the CT and MV CBCT. All steps (simulation, planning, setup verification, QA, and dose delivery) were performed in one 30 min session using phantoms. The monitor units (MU) for each plan were compared and dose distribution agreement was evaluated using mean dose difference over the entire volume and gamma index on the central 2D axial plane. All whole-brain dose distributions gave gamma passing rates higher than 95% for 2%/2 mm criteria, and pelvic sites ranged between 90% and 98% for 3%/3 mm criteria. However, thoracic spine treatments produced gamma passing rates as low as 47% for 3%/3 mm criteria. Our novel MV CBCT-based dose planning and delivery approach was feasible and time-efficient for the majority of cases. Limited MV CBCT FOV precluded workflow use for pelvic sites of larger patients and resulted in image clearance issues when tumor position was far off midline. The agreement of calculated MU on CT and MV CBCT was acceptable for all treatment sites.

Use of TrueBeam developer mode for imaging QA.

The purpose of this study was to automate regular Imaging QA procedures to become more efficient and accurate. Daily and monthly imaging QA for SRS and SBRT protocols were fully automated on a Varian linac. A three-step paradigm where the data are automatically acquired, processed, and analyzed was defined. XML scripts were written and used in developer mode in a TrueBeam linac to automatically acquire data. MATLAB R013B was used to develop an interface that could allow the data to be processed and analyzed. Hardware was developed that allowed the localization of several phantoms simultaneously on the couch. 14 KV CBCTs from the Emma phantom were obtained using a TrueBeam onboard imager as example of data acquisition and analysis. The images were acquired during two months. Artifacts were artificially introduced in the images during the reconstruction process using iTool reconstructor. Support vector machine algorithms to automatically identify each artifact were written using the Machine Learning MATLAB R2011 Toolbox. A daily imaging QA test could be performed by an experienced medical physicist in 14.3 ± 2.4 min. The same test, if automated using our paradigm, could be performed in 4.2 ± 0.7 min. In the same manner, a monthly imaging QA could be performed by a physicist in 70.7 ± 8.0 min and, if fully automated, in 21.8 ± 0.6 min. Additionally, quantitative data analysis could be automatically performed by Machine Learning Algorithms that could remove the subjectivity of data interpretation in the QA process. For instance, support vector machine algorithms could correctly identify beam hardening, rings and scatter artifacts. Traditional metrics, as well as metrics that describe texture, are needed for the classification. Modern linear accelerators are equipped with advanced 2D and 3D imaging capabilities that are used for patient alignment, substantially improving IGRT treatment accuracy. However, this extra complexity exponentially increases the number of QA tests needed. Using the new paradigm described above, not only the bare minimum ­ but also best practice ­ QA programs could be implemented with the same manpower.

Multi-modal fusion and feature enhancement U-Net coupling with stem cell niches proximity estimation for voxel-wise GBM recurrence prediction.

Objective.We aim to develop a Multi-modal Fusion and Feature Enhancement U-Net (MFFE U-Net) coupling with stem cell niche proximity estimation to improve voxel-wise Glioblastoma (GBM) recurrence prediction.Approach.57 patients with pre- and post-surgery magnetic resonance (MR) scans were retrospectively solicited from 4 databases. Post-surgery MR scans included two months before the clinical diagnosis of recurrence and the day of the radiologicaly confirmed recurrence. The recurrences were manually annotated on the T1ce. The high-risk recurrence region was first determined. Then, a sparse multi-modal feature fusion U-Net was developed. The 50 patients from 3 databases were divided into 70% training, 10% validation, and 20% testing. 7 patients from the 4th institution were used as external testing with transfer learning. Model performance was evaluated by recall, precision, F1-score, and Hausdorff Distance at the 95% percentile (HD95). The proposed MFFE U-Net was compared to the support vector machine (SVM) model and two state-of-the-art neural networks. An ablation study was performed.Main results.The MFFE U-Net achieved a precision of 0.79 ± 0.08, a recall of 0.85 ± 0.11, and an F1-score of 0.82 ± 0.09. Statistically significant improvement was observed when comparing MFFE U-Net with proximity estimation couple SVM (SVMPE), mU-Net, and Deeplabv3. The HD95 was 2.75 ± 0.44 mm and 3.91 ± 0.83 mm for the 10 patients used in the model construction and 7 patients used for external testing, respectively. The ablation test showed that all five MR sequences contributed to the performance of the final model, with T1ce contributing the most. Convergence analysis, time efficiency analysis, and visualization of the intermediate results further discovered the characteristics of the proposed method.Significance. We present an advanced MFFE learning framework, MFFE U-Net, for effective voxel-wise GBM recurrence prediction. MFFE U-Net performs significantly better than the state-of-the-art networks and can potentially guide early RT intervention of the disease recurrence.